> > Of course, drivers shouldn't be using f_pos outside their llseek > > method, as they should all behave the same with pread/pwrite as with > > llseek+read/write. > > Might not a driver update f_pos after read/write? The driver updates the passed pointer to an offset. It has no idea how to lock that and that is isolated and handled higher up the stack. There are no obvious reasons to peer at f_pos. I've so far checked all of drivers/char and that is clean (as well as being the most likely suspect for old code). A sweep of the other driver subsystems should be enough to find any offenders Alan -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
